INFO 4940/5940: Applied Machine Learning: Methods and Applications
This page contains an outline of the topics, content, and assignments for the semester. Note that this schedule will be updated as the semester progresses and the timeline of topics and assignments might be updated throughout the semester.
| WEEK | DATE | TOPIC | PREPARE | 
Prepare (R)
|
Prepare (Python)
|
MATERIALS | DUE |
|---|---|---|---|---|---|---|---|
| 1 | Tue, Aug 26 | Welcome to INFO 4940/5940 | π©βπ» Login to Cornellβs GitHub server |
π½οΈ slides 01 | |||
| Thu, Aug 28 | Case study in ML: Property assessment in Cook County | π Automated valuation model for all class 200 residential properties in Cook County Read the sections on Model Overview, Ongoing Issues, and FAQs |
π½οΈ slides 02 β¨οΈ ae 01 β¨οΈ hw 00 |
||||
| 2 | Tue, Sep 2 | Make a model | π isl - ch 2.1 |
π tmwr - ch 1-2, 4, 6 This will help you learn the {tidymodels} syntax. You can heavily skim chapters 1-2, 4. |
π pdsh - ch 5.2 - introducing Scikit-learn | π½οΈ slides 03 β¨οΈ ae 02 |
|
| Wed, Sep 3 | β
hw 00 - Python β hw 00 - R |
HW 00 by 11:59pm | |||||
| Thu, Sep 4 | Use your data | π isl - ch 5.1 π aml - ch 9-10 |
π tmwr - ch 5, 10 | π scikit-learn documentation - 3.1
|
π½οΈ slides 04 β¨οΈ ae 03 β¨οΈ hw 01 |
||
| 3 | Tue, Sep 9 | Build better training data | π aml - ch 6 | π tmwr - ch 8 | π pdsh - ch 5.5 - Feature engineering π scikit-learn documentation - 7.1
|
π½οΈ slides 05 β¨οΈ ae 04 |
|
| Wed, Sep 10 | β
hw 01 - Python β hw 01 - R |
HW 01 by 11:59pm | |||||
| Thu, Sep 11 | Build better training data | π aml - ch 5, 8.1-.3 | π½οΈ slides 06 β¨οΈ ae 05 β¨οΈ hw 02 |
||||
| 4 | Tue, Sep 16 | Tune your workflows | π aml - ch 11-12 - read ch 11 more closely, lightly skim ch 12 π isl - ch 8.2 - review of ensemble tree methods |
π tmwr - ch 12-14 | π scikit-learn documentation - 3.2
|
π½οΈ slides 07 β¨οΈ ae 06 |
|
| Wed, Sep 17 | β
hw 02 - Python β hw 02 - R |
HW 02 by 11:59pm | |||||
| Thu, Sep 18 | Evaluate models using appropriate metrics | π aml - ch 15 | π tmwr - ch 9 | π scikit-learn documentation - 3.4 - know how to implement different metrics at different stages of the ML process (e.g. resampling, evaluating test set performance). You donβt need to know every single metric defined in the documentation. |
π½οΈ slides 08 β¨οΈ ae 07 β¨οΈ hw 03 |
||
| 5 | Tue, Sep 23 | Feature selection | π isl - ch 6 - read sections 6.2-.3 | π½οΈ slides 09 β¨οΈ ae 08 |
|||
| Wed, Sep 24 | β
hw 03 - Python β hw 03 - R |
HW 03 by 11:59pm | |||||
| Thu, Sep 25 | Exploratory analysis | π r4ds - ch 10 - even if you use Python, a great overview of common EDA strategies (and easily transferrable to the plotnine implementation) |
π½οΈ slides 10 β¨οΈ ae 09 - Python β¨οΈ ae 09 - R β ae-09 |
||||
| 6 | Tue, Sep 30 | Project management | π§ Brainstorm a potential topic for the first project | π½οΈ slides 11 β¨οΈ ae 10 |
|||
| Thu, Oct 2 | Preprocess your data | π½οΈ slides 12 β¨οΈ ae 11 β¨οΈ hw 04 |
|||||
| 7 | Tue, Oct 7 | Explaining models through agnostic approaches | π iml - ch 2-3 | π½οΈ slides 13 β¨οΈ ae 12 |
|||
| Wed, Oct 8 | HW 04 by 11:59pm | ||||||
| Thu, Oct 9 | No class (out-of-town) | Project 01 proposal by 11:59pm | |||||
| 8 | Tue, Oct 14 | No class (Fall Break) | |||||
| Thu, Oct 16 | Document models | π βModel Cards for Model Reportingβ π Llama 3.2 documentation on HuggingFace |
π½οΈ slides 14 β¨οΈ ae 13 |
||||
| 9 | Tue, Oct 21 | Version and deploy models using APIs | π½οΈ slides 15 β¨οΈ ae 14 - Python β¨οΈ ae 14 - R β ae 14 |
||||
| Thu, Oct 23 | Publish APIs using Docker + cloud hosting | β¨οΈ If you are completing assignments on your local computer, install Docker Desktop | π½οΈ slides 16 β¨οΈ ae 15 - Python β¨οΈ ae 15 - R β ae 15 β¨οΈ hw 05 |
Project 01 EDA by 11:59pm | |||
| 10 | Tue, Oct 28 | An introduction to LLMs | π©βπ» Generate your API keys for OpenAI and Anthropic | π½οΈ slides 17 β¨οΈ ae 16 β ae 16 |
|||
| Wed, Oct 29 | HW 05 by 11:59pm | ||||||
| Thu, Oct 30 | Programming with LLMs | π½οΈ slides 18 β¨οΈ ae 17 β ae 17 |
Project 01 draft by 11:59pm | ||||
| 11 | Tue, Nov 4 | Prompt engineering | To be posted | To be posted | To be posted | ||
| Thu, Nov 6 | Augmented generation | To be posted | To be posted | To be posted | |||
| 12 | Tue, Nov 11 | Tool calling | To be posted | To be posted | To be posted | ||
| Thu, Nov 13 | Interactive dashboards powered by LLMs | To be posted | To be posted | To be posted | |||
| 13 | Tue, Nov 18 | An AI guided introduction to Shiny | To be posted | To be posted | To be posted | ||
| Thu, Nov 20 | Creating interactive chat bots with Shiny | To be posted | To be posted | To be posted | |||
| 14 | Tue, Nov 25 | To be posted | To be posted | To be posted | |||
| Thu, Nov 27 | No class (Thanksgiving Break) | To be posted | To be posted | To be posted | |||
| 15 | Tue, Dec 2 | To be posted | To be posted | To be posted | |||
| Thu, Dec 4 | Wrap-up: Where to go from here | To be posted | To be posted | To be posted | |||
| Thu, Dec 18 | Project 02 presentation | To be posted | To be posted | To be posted | Project 02 presentations at 12pm |